Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Look up keyword
Like this
0Activity
0 of .
Results for:
No results containing your search query
P. 1
1401.7251

1401.7251

Ratings: (0)|Views: 1|Likes:
Published by crocoali
aaa
aaa

More info:

Categories:Types, Brochures
Published by: crocoali on Jan 29, 2014
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

01/29/2014

pdf

text

original

 
Transparency of Strong Gravitational Waves
Y. Hadad, V. Zakharov
Department of Mathematics, University of Arizona, Tucson, Arizona, 85721 USA
(Dated: January 28th, 2014)This paper studies diagonal spacetime metrics. It is shown that the overdetermined Einsteinvacuum equations are compatible if one Killing vector exists. The stability of plane gravitationalwaves of the Robinson type is studied. This stability problem bares a fantastic mathematicalresemblance to the stability of the Schwarzschild black hole studied by Regge and Wheeler. Justlike for the Schwarzschild black hole, the Robinson gravitational waves are proven to be stable withrespect to small perturbations. We conjecture that a bigger class of vacuum solutions are stable,among which are all gravitational solitons. Moreover, the stability analysis reveals a surprising fact:a wave barrier will be transparent to the Robinson waves, which therefore passes through the barrierfreely. This is a hint of integrability of the 1 + 2 vacuum Einstein equations for diagonal metrics.
PACS numbers: 02.30.Ik,02.30.Jr,05.45.Yv,04.20.-q,04.30.-w,04.70.Bw
I. INTRODUCTION
In the theory of relativity, the Einstein-Hilbert actionis
 = 12
 
 R
√ 
gd
4
x
 (1)where
 R
 is the scalar curvature of the spacetime metric
g
µν 
,
 g
 is the determinant of 
 g
µν 
 and the integration isperformed over the four-dimensional spacetime. Varyingthe Einstein-Hilbert action (1) with respect to the inversemetric
 g
µν 
gives Einstein’s vacuum equations,
R
µν 
 = 0 (2)where
 R
µν 
 is the Ricci curvature tensor. Einstein’s vac-uum equations determine the evolution of the spacetimemetric
 g
µν 
 in empty space.This paper focuses on
 diagonal 
 spacetime metrics.These are metrics that can be written in the form
g
µν 
 = (
µ
)
2
δ 
µν 
 (3)where
 δ 
µν 
 is the Kroncker delta. Here and in the rest of this paper, Einstein’s summation convention is
 not 
 used.In matrix form, the diagonal metric is
g
µν 
 =
(
0
)
2
0 0 00 (
1
)
2
0 00 0 (
2
)
2
00 0 0 (
3
)
2
.
 (4)It is convenient not to worry about the sign of the metric.Instead, one may restore the proper metric signature (
+++) by substituting
 H 
0
 →
iH 
0
.It is a well known result that
 every 
 metric
 g
µν 
 may bediagonalized at
 any 
 given event of spacetime (e.g. by us-ing Riemann normal coordinates) [1]. Nevertheless, thisis a local result, which holds globally only for very specificspacetime metrics. This means that the class of metricsthat can be casted into the diagonal form (3)
 globally 
should be expected to have
 unique features 
. It is impor-tant to keep in mind that the
 diagonality 
 of the metric
is not an invariant property 
. In other words, some non-diagonal metrics
 g
µν 
 may be transformed to the diagonalform (3) by a proper choice of coordinates.The metric (3) describes a wide range of physical phe-nomena. In particular, it includes the Schwarzschildblack hole [2], the Kasner metric [3], the Friedmann- Robertson-Walker model of cosmology [4], the Milnemodel of cosmology [4], a certain class of single-polarizedplane gravitational waves [4] and special cases of gravi-tational solitons [5].The goal of the rest of this paper is to study the systemof vacuum Einstein Eqs. (2) for diagonal metrics (3). In particular, the mathematical structure of the equationsand their physical implications on gravitational waves areemphasized.Section II includes a derivation of the Einstein equa-tions in the case of the diagonal metric (3). A convenientform for analyzing the equations is obtained. Section IIIshows that if at least one Killing vector exists, Einstein’sequations for diagonal metric are compatible. In sectionIV, plane gravitational waves are studied. A simple crite-ria for asymptotic flatness and compatibility of the fieldequations for plane waves are derived. One of the mostfamous examples of such plane waves is the Bondi-Pirani-Robinson (BPR) waves [6]. In section V it is proven that such waves are stable with respect to diagonal perturba-tions that depend on 1 + 2 coordinates.As a concrete example, in section VI a BPR wave withsoliton-like properties is studied. The emitted (pertur-bation) wave is shown to travel through the BPR wave
without any reflection 
 and
 independently of the amplitude of the BPR wave 
. The latter implies that a
 strong (BPR)gravitational wave would be transparent 
 to the perturba-tion wave. The only remnant of the collision is a phaseshift which depends on the angle between the two waves.These properties, which are typically exhibited by soli-tons,
 suggest that the 
 1 + 2
 vacuum Einstein equations  for diagonal metrics are integrable 
, similarly to the 1+1vacuum Einstein equations [7, 8].
  a  r   X   i  v  :   1   4   0   1 .   7   2   5   1  v   1   [  g  r  -  q  c   ]   2   8   J  a  n   2   0   1   4
 
2
II. THE FIELD EQUATIONS
For the diagonal metric (3), the inverse metric is
g
µν 
= 1(
µ
)
2
δ 
µν 
(5)and the Christoffel symbols areΓ
λµν 
 = 0 (6)Γ
µµν 
 =
 ∂ 
ν 
 (ln
µ
)Γ
ν µµ
 =
 
 1(
ν 
)
2
µ
∂ 
ν 
µ
where
 µ,ν,λ
 are assumed to be mutually exclusive indices(
µ
=
 ν 
,
 µ
=
 λ
,
 ν 
 
=
 λ
). Define the rotation coefficients
Q
µν 
 = 1
ν 
∂ 
ν 
µ
 (7)with which one can write the off-diagonal Ricci curvaturetensor as
R
µν 
 =
λ
=
µ,ν 
µ
λ
(
∂ 
ν 
Q
λµ
Q
λν 
Q
νµ
) (8)for
 µ
=
 ν 
. As for the diagonal elements, the Ricci tensorgives
R
µµ
 =
ν 
=
µ
µ
ν 
µν 
 (9)where
µν 
 =
 ∂ 
ν 
Q
µν 
 +
 ∂ 
µ
Q
νµ
 +
λ
=
µ,ν 
Q
µλ
Q
νλ
 (10)The scalar curvature is
R
 =
2
µ<ν 
µν 
µ
ν 
.
 (11)Since the determinant of the metric is det
g
 =(
0
1
2
3
)
2
, the Einstein-Hilbert action (1) is
 =
i
µ
=
ν 
=
λ
=
σ
 
 
µν 
λ
σ
d
4
x.
 (12)If one performs an integration by parts to remove thesecond order derivatives in
 
µν 
, a very concise formulafor the Einstein-Hilbert action in terms of the metric co-efficients only is obtained,
 =
µ
 
 i
µ
ν 
=
λ
=
σ
=
µ
ν 
(
∂ 
µ
λ
)(
∂ 
µ
σ
)
d
4
x.
 (13)The reader should not be alarmed by the appearanceof the imaginary root of unity
 i
 =
√ 
1. It is there dueto the signature of the metric and the transformation
0
 →
 iH 
0
, which was mentioned after Eq. (4), revealsimmediately that the action (13) is manifestly real-valuedas expected.
III. COMPATIBILITY
When studying
 general 
 metrics, the symmetric
 g
µν 
 hasten elements, four of which may be eliminated throughthe use of gauge transformations. This makes the vacuumEinstein equations
 R
µν 
 = 0 an overdetermined system of ten equations for six unknowns. In normal circumstancethis might raise the question of compatibility. Neverthe-less, this is not an issue, as one can prove using the fourBianchi identities that the vacuum Einstein equations areindeed compatible [1].However, the situation is rather different when dis-cussing diagonal metrics (3). In this case, Einstein’s vac-uum equations
 R
µν 
 = 0 give ten equations again, but thistime for only
 four 
 unknown functions
 H 
µ
 (
µ
 = 0
,
1
,
2
,
3).In this case, the usual argument using the Bianchi identi-ties ceases to hold, and an important question thus arises:
are the Einstein’s vacuum equations for diagonal metrics compatible? 
Consider the diagonal metric in Eq. (3) with the ad-ditional assumption that it is independent of 
 x
3
. Math-ematically, this means that the metric has the Killingvector
 ∂ 
3
 and depends on the three coordinates
 x
0
,x
1
,x
2
only. In this case, the off-diagonal terms of the Ricci cur-vature tensor (8) give only three independent equations
R
01
 =
 R
02
 =
 R
12
 = 0 (14)coupled to the four diagonal equations
R
µµ
 = 0 (15)for
 µ
 = 0
,
1
,
2
,
3. Eqs. (14) and (15) will be referred to as
 the 
 1+2
 vacuum Einstein equations for diagonal met-rics 
. The 1 + 2 vacuum Einstein equations for diagonalmetrics form an overdetermined system of seven equa-tions for four unknown functions. As an overdeterminedsystem, the compatibility of the seven equations mustbe proven, as it does not follow from the argument typ-ically used for non-diagonal metrics. The authors couldnot find any evidence for such a result in the literature.Whether the 1+2 Einstein equations for diagonal metricsare indeed compatible is a very natural question to ask,as such metrics have many applications in cosmology andastronomy, some of which will be described in the nextsections. Fortunately, it turns out that the answer is af-firmative, as the next theorem proves.
Theorem 16.
 The 
 1+2
 Einstein equations for diagonal metrics ( 14,15 ) are compatible.Proof.
 Proving the statement of this theorem using theoriginal degrees of freedom
 H 
0
,
1
,
2
 and
 
3
 is rathertedious. Instead, it is much easier to exploit the specialrole of 
 
3
 as the degree of freedom that corresponds tothe Killing vector
 ∂ 
3
. Define,
0
 =
 e
Λ
γ
1
 =
 e
Λ
β 
 (17)
2
 =
 e
Λ
α
3
 =
 e
Λ
.
 
3Using the new degrees of freedom
 α,β,γ 
 and Λ, the off-diagonal Einstein Eqs. (14) are(18)
∂ 
0
∂ 
1
α
 =
 
2
α
(
∂ 
0
Λ)(
∂ 
1
Λ) + (
∂ 
0
β 
)(
∂ 
1
α
)
β 
 + (
∂ 
0
α
)(
∂ 
1
γ 
)
γ ∂ 
0
∂ 
2
β 
 =
 
2
β 
(
∂ 
0
Λ)(
∂ 
2
Λ) + (
∂ 
0
α
)(
∂ 
2
β 
)
α
 + (
∂ 
0
β 
)(
∂ 
2
γ 
)
γ ∂ 
1
∂ 
2
γ 
 =
 
2
γ 
(
∂ 
1
Λ)(
∂ 
2
Λ) + (
∂ 
1
α
)(
∂ 
2
γ 
)
α
 + (
∂ 
1
γ 
)(
∂ 
2
β 
)
β  .
As for the diagonal Eqs. (15), it is convenient to repre-sent them in an equivalent form through the variationalformulation. The Lagrangian density of the Einstein-Hilbert action (13) is now
L
 = 2
αβ γ 
 (
∂ 
0
Λ)
2
 αγ β 
 (
∂ 
1
Λ)
2
 βγ α
 (
∂ 
2
Λ)
2
(19)
(
∂ 
0
α
)(
∂ 
0
β 
)
γ 
 + (
∂ 
1
α
)(
∂ 
1
γ 
)
β 
 + (
∂ 
2
β 
)(
∂ 
2
γ 
)
α
.
The variations
 δS δα
 =
 δS δβ
 =
 δS δγ 
 = 0 give three of thediagonal equations
β∂ 
0
∂ 
0
β 
γ∂ 
1
∂ 
1
γ 
 =
 
β 
2
(
∂ 
0
Λ)
2
+
 γ 
2
(
∂ 
1
Λ)
2
 β 
2
γ 
2
α
2
 (
∂ 
2
Λ)
2
+
 β γ 
(
∂ 
0
β 
)(
∂ 
0
γ 
)
 γ β 
(
∂ 
1
β 
)(
∂ 
1
γ 
) +
 βγ α
2
(
∂ 
2
β 
)(
∂ 
2
γ 
) (20)
α∂ 
0
∂ 
0
α
γ∂ 
2
∂ 
2
γ 
 =
 
α
2
(
∂ 
0
Λ)
2
 α
2
γ 
2
β 
2
 (
∂ 
1
Λ)
2
+
 γ 
2
(
∂ 
2
Λ)
2
+
 αγ 
(
∂ 
0
α
)(
∂ 
0
γ 
) +
 αγ β 
2
 (
∂ 
1
α
)(
∂ 
1
γ 
)
 γ α
(
∂ 
2
α
)(
∂ 
2
γ 
)
α∂ 
1
∂ 
1
α
 +
 β∂ 
2
∂ 
2
β 
 =
 
α
2
β 
2
γ 
2
 (
∂ 
0
Λ)
2
α
2
(
∂ 
1
Λ)
2
β 
2
(
∂ 
2
Λ)
2
+
 αβ γ 
2
 (
∂ 
0
α
)(
∂ 
0
β 
) +
 αβ 
(
∂ 
1
α
)(
∂ 
1
β 
) +
 β α
(
∂ 
2
α
)(
∂ 
2
β 
)
,
while the last diagonal equation,
 δS δ
Λ
 = 0 is
∂ 
0
αβ γ  ∂ 
0
Λ
∂ 
1
αγ β  ∂ 
1
Λ
∂ 
2
βγ α ∂ 
2
Λ
 = 0
.
 (21)To prove the statement of the theorem, we differentiateeach of Eqs. (20) with respect to
 x
2
,x
1
and
 x
0
respec-tively. This gives three third order equations for
 α,β 
 and
γ 
. One may now eliminate each of the third order termsusing the non-diagonal Eqs. (18). After a lengthy alge-bra, one sees that with the aid of Eqs. (18) once more,all 38 terms in each equation completely vanish. There-fore the 1+2 Einstein equations for diagonal metrics areindeed compatible.The degrees of freedom
 α,β,γ 
, and Λ from theorem(16) are very useful. They provide an alternative way tostudy general diagonal spacetime metrics (3). With suchdegrees of freedom the spacetime interval is
ds
2
=
 e
(
γdx
0
)
2
+ (
βdx
1
)
2
+ (
αdx
2
)
2
+
e
(
dx
3
)
3
(22)The spacetime interval (22) naturally generalizes theinterval studied in [7, 8]. To see this, assume the metric is independent of 
 x
2
, set
 β 
 =
 γ 
 and define
 
 =
 γ 
2
e
.This turns the spacetime interval (22) into
ds
2
=
 f 
(
dx
0
)
2
+ (
dx
1
)
2
+
 α
2
e
(
dx
2
)
2
+
 e
(
dx
3
)
3
(23)For this metric one may use the inverse scattering trans-form [7, 8] to derive gravitational solitons on diagonalmetrics [5].There is another merit of using the new degrees of free-dom, as in the course of the proof of theorem (16) we just derived
 a new conservation law 
. This is of courseEq. (21). Whenever the metric is asymptotically flat, italso yields the integral of motion,
 =
 
αβ γ  ∂ 
0
Λ
dx
1
dx
2
(24)which is the conjugate momentum of the function Λ, ascan be easily seen from the Lagrangian in Eq. (19).
IV. PLANE GRAVITATIONAL WAVES
Waves come in many forms and shapes. The simplestof which are of course plane waves, whose wavefronts areparallel planes extended ad infinitum. In general rela-tivity, plane gravitational waves are typically studied asa special case of the famous
 pp
-waves [9]. The
 pp
-classconsists of any spacetime metric that can be casted intothe form,
ds
2
=
 H 
(
u,x,y
)
du
2
+ 2
dudv
 +
 dx
2
+
 dy
2
.
 (25)Recently, a coordinate-free definition of them was given[10]. For such a metric, Einstein’s vacuum equation re-duces to Laplace’s equation,
∂ 
2
∂x
2
 +
 ∂ 
2
∂y
2
 = 0 (26)and is therefore
 linear 
 in
 H 
. A
 pp
-wave is called a
 plane wave 
 if 
 H 
 can be transformed into
(
u,x,y
) =
 a
(
u
)(
x
2
y
2
) + 2
b
(
u
)
xy
 (27)

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->